For our client we have a website running in Netlify (Nuxt 3.14). We use the Umbraco Content Delivery API to get content and populate the site with the content that the client has configured in Umbraco.
We’re using Umbraco 15.3.1 and we ran into an issue last Sunday.
From Sunday midnight we suddenly received no navigation items from the content delivery api anymore, while just before midnight it did return the appropriate data.
Has anyone seen this before, where the content delivery api suddenly didn’t return any items anymore?
Interesting fact is that there were no logs being written due to storage issues on the environment. These storage issues were already present for a week and we didn’t notice it (the website and CMS performed fine and no notification was send that storage was full).
We already reached out to Umbraco Support regarding this, but so far no luck on that front.
First thing I did was restart the CMS and somehow that fixed the issue at first. The navigation items were being returned through the Content Delivery API.
Still, it was weird that it worked again, because there still was no storage space left on disk.
For some reason, the internal index folder on disk was 6GB (still not sure why) and caused the logs and the indexes to not rebuild at all.
I have not tried to rebuild the memory cache at that moment. Would that maybe fix the internal index from being 6GB?
I don’t think the memory cache has anything to do with the index. 6GB index is pretty large, I randomly checked one of our production sites and the internal index is 55MB. It depends on the number of nodes ofcourse, but getting to 6GB would be massive in that case.
On what platform are you hosting? And you could just stop the Umbraco instance, delete the indexes from disk and start again.
We have restarted the instance from the Umbraco Cloud Portal, but that didn’t fix the index. The only scenario that I haven’t tried is deleting the indexes and immediately after restarting the CMS.
As far as I know there’s no way to stop one of the environment entirely from the Umbraco Cloud Portal.
We have a lot of items, but not as big as 6 GB. At the moment, the index folder is stable and around 14 MB.
@Luuk
They are looking into it, but it seems like it’s a tough nut to crack.
For now it’s less important, as the platform is stable again. I just want to get to the bottom of it why this happens and how it can be prevented in the future.
